Efficient placement of Software-Defined Networking (SDN) controllers in vehicular networks is crucial to optimize communication performance, ensure load balancing, and achieve fault tolerance. This paper presents a novel multi-objective optimization approach MOTA-SVB for SDN controller placement in vehicular networks, leveraging vehicle trajectory prediction to accommodate mobility. A trajectory-aware placement strategy is proposed to minimize the predicted physical distance between vehicles and Roadside Units (RSUs), reduce communication delay, and ensure resilience using Byzantine Fault Tolerance (BFT). To address the trade-offs between conflicting objectives, we employ the novel S-Metric Selection Evolutionary Multi-Objective Algorithm (SMS-EMOA), which balances the minimization of vehicle-to-RSU distance, number of deployed controllers, and communication delay, while satisfying load balancing constraints based on vehicle-to-controller mappings. Extensive simulations demonstrate that the proposed approach outperforms the state-of-the-art NBFT-SDN by achieving up to a 21% reduction in communication delay, about 47% improvement in load balancing, and a 25–30% reduction in the number of deployed controllers under varying failure scenarios.
Haq et al. (Mon,) studied this question.